Hyperspectral Image Restoration and Super-resolution with Physics-Aware Deep Learning for Biomedical Applications
Authors:
Yuchen Xiang,
Zhaolu Liu,
Monica Emili Garcia-Segura,
Daniel Simon,
Boxuan Cao,
Vincen Wu,
Kenneth Robinson,
Yu Wang,
Ronan Battle,
Robert T. Murray,
Xavier Altafaj,
Luca Peruzzotti-Jametti,
Zoltan Takats
Abstract:
Hyperspectral imaging is a powerful bioimaging tool which can uncover novel insights, thanks to its sensitivity to the intrinsic properties of materials. However, this enhanced contrast comes at the cost of system complexity, constrained by an inherent trade-off between spatial resolution, spectral resolution, and imaging speed. To overcome this limitation, we present a deep learning-based approac…
▽ More
Hyperspectral imaging is a powerful bioimaging tool which can uncover novel insights, thanks to its sensitivity to the intrinsic properties of materials. However, this enhanced contrast comes at the cost of system complexity, constrained by an inherent trade-off between spatial resolution, spectral resolution, and imaging speed. To overcome this limitation, we present a deep learning-based approach that restores and enhances pixel resolution post-acquisition without any a priori knowledge. Fine-tuned using metrics aligned with the imaging model, our physics-aware method achieves a 16X pixel super-resolution enhancement and a 12X imaging speedup without the need of additional training data for transfer learning. Applied to both synthetic and experimental data from five different sample types, we demonstrate that the model preserves biological integrity, ensuring no features are lost or hallucinated. We also concretely demonstrate the model's ability to reveal disease-associated metabolic changes in Downs syndrome that would otherwise remain undetectable. Furthermore, we provide physical insights into the inner workings of the model, paving the way for future refinements that could potentially surpass instrumental limits in an explainable manner. All methods are available as open-source software on GitHub.
△ Less
Submitted 3 March, 2025;
originally announced March 2025.
Multi-modal Heart Failure Risk Estimation based on Short ECG and Sampled Long-Term HRV
Authors:
Sergio González,
Abel Ko-Chun Yi,
Wan-Ting Hsieh,
Wei-Chao Chen,
Chun-Li Wang,
Victor Chien-Chia Wu,
Shang-Hung Chang
Abstract:
Cardiovascular diseases, including Heart Failure (HF), remain a leading global cause of mortality, often evading early detection. In this context, accessible and effective risk assessment is indispensable. Traditional approaches rely on resource-intensive diagnostic tests, typically administered after the onset of symptoms. The widespread availability of electrocardiogram (ECG) technology and the…
▽ More
Cardiovascular diseases, including Heart Failure (HF), remain a leading global cause of mortality, often evading early detection. In this context, accessible and effective risk assessment is indispensable. Traditional approaches rely on resource-intensive diagnostic tests, typically administered after the onset of symptoms. The widespread availability of electrocardiogram (ECG) technology and the power of Machine Learning are emerging as viable alternatives within smart healthcare. In this paper, we propose several multi-modal approaches that combine 30-second ECG recordings and approximate long-term Heart Rate Variability (HRV) data to estimate the risk of HF hospitalization. We introduce two survival models: an XGBoost model with Accelerated Failure Time (AFT) incorporating comprehensive ECG features and a ResNet model that learns from the raw ECG. We extend these with our novel long-term HRVs extracted from the combination of ultra-short-term beat-to-beat measurements taken over the day. To capture their temporal dynamics, we propose a survival model comprising ResNet and Transformer architectures (TFM-ResNet). Our experiments demonstrate high model performance for HF risk assessment with a concordance index of 0.8537 compared to 14 survival models and competitive discrimination power on various external ECG datasets. After transferability tests with Apple Watch data, our approach implemented in the myHeartScore App offers cost-effective and highly accessible HF risk assessment, contributing to its prevention and management.
△ Less
Submitted 29 February, 2024;
originally announced March 2024.
